HAL Id: hal-00660734
https://hal.archives-ouvertes.fr/hal-00660734
Submitted on 17 Jan 2012
HAL is a multi-disciplinary open access
archive for the deposit and dissemination of
sci-entific research documents, whether they are
pub-lished or not. The documents may come from
teaching and research institutions in France or
abroad, or from public or private research centers.
L’archive ouverte pluridisciplinaire HAL, est
destinée au dépôt et à la diffusion de documents
scientifiques de niveau recherche, publiés ou non,
émanant des établissements d’enseignement et de
recherche français ou étrangers, des laboratoires
publics ou privés.
The Properties of Star-forming Regions within a Galaxy
at Redshift 2
M. Swinbank, A. Edge, J. Richard, I. Smail, C. de Breuck, A. Lundgren, G.
Siringo, A. Weiss, A. Harris, A. Baker, et al.
To cite this version:
M. Swinbank, A. Edge, J. Richard, I. Smail, C. de Breuck, et al.. The Properties of Star-forming
Regions within a Galaxy at Redshift 2. The Messenger, 2010, 139, pp.42-45. �hal-00660734�
The Properties of Star-forming Regions within a Galaxy at
Redshift 2
Astronomical Science Mark Swinbank1 Alastair Edge1 Johan Richard1 Ian Smail1 Carlos De Breuck2 Andreas Lundgren2 Giorgio Siringo2 Axel Weiss3 Andrew Harris4 Andrew Baker5 Steve Longmore6 Rob Ivison71 Institute for Computational Cosmology,
Department of Physics, University of Durham, United Kingdom
2 ESO
3 Max-Planck-Institut für
Radioastrono-mie, Bonn, Germany
4 Department of Astronomy, University of
Maryland, College Park, USA
5 Department of Physics and Astronomy,
Rutgers University, Piscataway, USA
6 Harvard-Smithsonian Center for
Astro-physics, Cambridge, USA
7 United Kingdom Astronomy Technology
Centre, Edinburgh, Scotland
The discovery and subsequent follow-up of one of the brightest sub-mm gal-axies discovered so far is presented. First identified with the LABOCA instru-ment on APEX in May 2009, this galaxy lies at z = 2.32 and its brightness of 106 mJy at 870 µm is due to the gravita-tional magnification caused by a mas-sive intervening galaxy cluster. Follow-up observations with APEX SABOCA have been used to constrain the far-infrared spectral energy distribution and hence measure the star formation rate, and Swedish Heterodyne Facility Instru-ment observations help constrain the excitation of the cold molecular gas. Furthermore, high resolution follow-up with the sub-mm array resolves the star-forming regions on scales of just 100 parsecs. These results allow study of galaxy formation and evolution at a level of detail never before possible and provide a glimpse of the exciting pos-sibilities for future studies of galaxies at these early times, particularly with ALMA.
Massive galaxies in the early Universe have been shown to be forming stars at surprisingly high rates. The most prom-inent examples are sub-millimetre galax-ies (SMGs) whose star formation rates can exceed 1000 MA year–1 (Chapman
et al., 2003; Coppin et al., 2008). With this intense star formation rate, a massive galaxy (comparable to the stellar mass of a local elliptical galaxy) can be built in just 100 million years (Tacconi et al., 2008). As such, the sub-mm galaxy pop-ulation may represent the formation epoch of today’s massive elliptical galax-ies (Lilly et al., 1999; Smail et al., 1997). This is, however, a theoretically provoca-tive conclusion as it is at odds with theo-retical models. Indeed, sophisticated gal-axy formation models have had to alter their prescriptions for starbursts radically, invoking exotic physics such as strong variations in the stellar initial mass func-tion in order to account for the SMG pop-ulation (Baugh et al., 2004). Although controversial, there is growing evidence that the increased interstellar medium (ISM) pressure within the warm, dense gas in local ultra-luminous infrared galax-ies (ULIRGs) results in an increased Jeans mass (Perez-Torres et al., 2009). In order to test these controversial prescrip-tions, direct observational constraints on the properties of individual star-forming regions within high redshift galaxies are necessary. Such observations are technologically challenging, requiring the increased collecting area and sensitivity of the next generation tele scopes, such as the Extremely Large Tele scopes currently being planned and the Atacama Large Millimeter/Submillimeter Array (ALMA) under construction. Moreover, since the most intensely star-forming gal-axies are also the most obscured, they are optically faint and difficult to identify and study on sub-kiloparsec (sub-kpc) scales.
Gravitational lensing
The most promising route for investigat-ing the properties of high redshift galax-ies on sub-kpc scales is to use massive galaxy clusters as natural lenses. Galaxy clusters magnify the images of distant galaxies that serendipitously lie behind them. This natural magnification causes
background galaxies to be strongly amplified and stretched, providing us with the opportunity of studying young and intrinsically faint galaxies with a spatial resolution that cannot be attained via conventional observations. The natural amplification caused by the galaxy cluster has two effects: (i) the image of the back-ground galaxy is magnified at a fixed sur-face brightness (i.e. the total brightness is increased); and, (ii) the galaxy is not only amplified, but it is also stretched, mak- ing it possible to spatially resolve compo-nents of the galaxy from the ground.
LABOCA imaging of distant clusters
During recent Atacama Pathfinder Ex-periment (APEX) 870-µm observations of a massive galaxy cluster (MACS 2135-0102) we serendipitously discovered an extremely bright sub-mm galaxy with an 870-µm flux of 106 mJy (see Figure 1, left). This is three times brighter than any other known high-redshift star-forming galaxy and even brighter than the Clover-leaf and APM08279 quasars. Due to the high significance of the source, we were able to centroid the sub-mm emission to an accuracy of less than 0.5 arcseconds and identify the mid-infrared counterpart with a bright source detected by the Spitzer Space Telescope Infra-Red Array Camera (IRAC) and Multiband Imaging Photometer (MIPS); see Figure 1 right. Due to the brightness, we were able to obtain a redshift measurement for the galaxy of z = 2.3259 via the blind detec-tion of CO(1–0) using the Zpectrometer on the Green Bank Telescope (GBT), confirming that the galaxy is high redshift and strongly gravitationally lensed. Indeed, using a detailed mass model of the galaxy cluster we derive an amplifica-tion factor of 32.5 ± 4.5. Intrinsically therefore the 870-µm flux is ~ 3 mJy, typi-cal of the sub-mm background, and, on account of the amplification, an ideal tar-get for study. The CO(1–0) emission line flux can also be used to estimate the cold molecular gas mass; after correcting for lensing amplification we derive a gas mass of 2 × 1010 M
SABOCA
With confirmation that this is a highly am-plified background galaxy, observations at 350 µm using the SABOCA camera (see article by Siringo et al., p. 20) were carried out to parameterise the spectral energy distribution (SED); see Figure 2. With a measured 350-µm flux of 530 mJy we derive an intrinsic far-infrared lumi-nosity of 1.2 × 1012 L
A. This suggests a
star formation rate of 210 MA/yr. If this
star formation rate has been maintained, then it takes just ~ 120 Myr to build the observed stellar mass of 4 × 1010 M
A and
the remaining gas depletion timescale is a further 70 Myr. Together, this sug-gests that this intense star formation epi-sode may be the first major growth phase of this galaxy.
Figure 1. True colour HST VI-band image of the mas-sive galaxy cluster MACSJ2135-0102 (zcl = 0.325)
is shown (left). The white contours denote the APEX/ LABOCA 870-µm contours from the galaxy (con-tours denote 5, 10, 15, 20, 25, 30σ), identifying an SMG with flux 106mJy. The solid red lines denote the z = 2.326 radial and tangential critical curves from the best-fit lens model. The true colour IRAC 3.6, 4.5, 8.0-µm image of the cluster core is shown (right). The contours denote the intensity of the 350-µm map and show 5, 10, 15, 20σ. The red lines again show the lens model critical curves.
@Q BR DB N MC R @QBRDBNMCR D σ σ σ σ σ σ σ σ σ σ σ σ l l l l l @QBRDBNMCR l l l E %K TW C DM RH SX L )X 6@UDKDMFSG§L Y +ANK + , QO 3C * 3 C *
Figure 2. Spectral energy distribution (SED) of the lensed galaxy SMMJ2135-0102 is shown. To test how the SED of the galaxy compares to local star-bursts, the SEDs of M82 (blue dashed line) and Arp220 (red dashed line), both arbitrarily scaled to the 1.2 mm flux, are overlaid. The solid black line
denotes the best fit modified blackbody to the pho-tometry. Accounting for lensing amplification, the bolometric luminosity of the galaxy is Lbol = 1.2 ± 0.2
× 1012 L
A which corresponds to a star formation rate
Molecular gas properties
To further constrain the properties of the cold molecular gas, observations of the high-J CO emission (up to J = 7) were obtained with a combination of Institut de Radioastronomie Millimétrique Plateau de Bure Interferometer (IRAM PdBI; CO(3–2) up to CO(7–6)) and APEX Swedish Heter-odyne Facility Instrument (SHFI) (CO(7–6)). In Figure 3a we show the CO(1–0) and CO(7–6) line profiles obtained with the IRAM Eight MIxer Receiver (EMIR) and APEX/SHFI respectively. All the CO spec-tra show multiple peaks that we decom-pose into a “red” component and a “blue” component which are separated by 300 ± 24 km/s. Since the ratio of the high/ low-J emission lines reflects the excita-tion of the molecular gas, we extract and model the CO spectral line energy distri-bution (SLED) of both “red” and “blue” components using large velocity gradient (LVG) models (e.g., Weiss et al., 2005). Both components are best fit with two excitation models (with temperatures ~ 40 K and densities of approximately 103
and 104 cm–3; see Figure 3b. The LVG
modelling also includes the “equivalent size” of the CO-emitting region as a free parameter, and we are therefore able to estimate the size of each component indirectly. The SLED modelling suggests that both the “red” and “blue” compo-nents appear to have a compact core with radius ~ 150 pc surrounded by a more extended, diffuse region with radius ~ 500 pc. Although these equivalent sizes come with some uncertainty, the brightest CO-emitting regions appear similar to those seen locally as starbursts with a dense, active star-forming core surrounded by a diffuse gas disc, per-haps providing the first direct evidence for a diffuse disc in an SMG.
Resolving the star-forming regions
To resolve the sub-mm emission directly, Submillimeter Array (SMA) observations were used to image the galaxy’s 870-µm (345 GHz) continuum emission with a 0.2-arcsecond synthesised beam and hence investigate the sub-mm morphol-ogy. The map resolves the galaxy into eight discrete components over ~ 4 arc-seconds in projection (Figure 4). These components represent two mirror images
of the galaxy each comprising four sepa-rate emission regions. Reconstructing the source-plane image using the lensing model, the galaxy comprises four bright star-forming regions in the source plane (indicated as A, B, C and D in Figure 4), that are separated by 1.5 kpc in projection. At the most highly amplified posi -tion, the source-plane resolution reaches ~ 90 pc which is comparable to the char-acteristic size of giant molecular clouds in the Milky Way (Scoville et al., 1989). We compare the sizes and luminosi- ties of the star-forming regions within SMMJ2135-0102 to giant molecular clouds (GMCs) in the local Universe (Scoville et al., 1989; Snell et al., 2002; Caldwell et al., 1996) in Figure 5. With-
in typical GMCs, constant energy density leads to a correlation between size and 260-µm luminosity, such that luminosity is proportional to radius cubed. However, within the dense cores of GMCs and young Hii regions, massive stars
domi-nate the emission and produce luminosity densities a factor ~ 100 times higher than in typical GMCs (Hill et al., 2005) and shown by the upper dashed line in Fig-ure 5. The star-forming regions within SMMJ2135-0102 are ~ 100 pc across, which is 100 times larger than dense GMC cores, but as Figure 5 shows their luminosities are approximately 100 times higher than expected for typical star-forming regions. Indeed, the luminosity densities of the star-forming regions within SMMJ2135-0102 are comparable Swinbank M. et al., The Properties of Star-forming Regions within a Galaxy at z = 2 Astronomical Science %K TW L )X %QDPTDMBX&'Y 5DKNBHSXJLR ".l l l l (BN) ) X JL R !KTDBNLONMDMS )TOODQ 5DKNBHSXJLR %QDPTDMBX&'Y ".l l l l (BN) " . l 1DCBNLONMDMS )TOODQ
Figure 3a (above). Spectra of the 12CO(1–0) and 12CO(7–6) emission from SMMJ2135-0102 obtained
with GBT/Zpec (left) and APEX/SFHI (right). The CO emission shows at least two velocity components separated by 300 km/s.
Figure 3b (below). The spectral line energy distribu-tions (SLEDs) for the “blue” and “red” velocity com-ponents seen in CO emission.
2,,) QO &H@MS,NKDBTK@Q"KNTCR 8NTMF'((QDFHNMR #DMRD&,"BNQDR + ¨L 6 ' Y 2HYDOB @QBRDBNMCR @Q BR DB N MC R l l l l l l $ % & ' ' & % $ 6RXUFHSODQH QJOB Q J O B l l l l l l $ % & '
Figure 5. Correlation between size and luminosity for star-forming regions in SMMJ2135-0102 compared to those in the Milky Way and local galaxies. Black squares denote the size and 260-µm luminosities of giant molecular clouds in the Milky Way and the local group. The lower dashed line shows a fit to this relation (with slope fixed at L260 µ ∝ r3; i.e. constant
energy density). The solid red points denote the sizes and luminosities of the star-forming regions in SMMJ2135-0102. The sizes and luminosities of dense cores of galactic GMCs in Henize 2-10 and M82 are plotted (open diamonds) as well as the two dominant star-forming regions within the local ultra-luminous infrared galaxy Arp220 (blue triangles). Figure 4a. The SMA map of SMMJ2135-0102 show-ing the image-plane morphology of the lensed galaxy at 870 µm. The galaxy comprises eight individual components, separated by up to 4 arcseconds in projection. The z = 2.326 radial critical curve (red line) is overlaid. The beam size is shown (lower right).
Figure 4b. The spatial distribution of the four compo-nents in the image plane from the lens model are shown. Each of the components (A, B, C and D), which are mirrored about the lensing critical curve, are identified.
to dense GMC cores, but with lumino-sities 107 times larger. Thus, it is likely
that each of the star-forming regions in SMMJ2135-0102 comprises ~ 107 dense
GMC cores.
The luminosity (and therefore star forma-tion) density of the star-forming regions within SMMJ2135-0102 are also similar to those found in the highest density re-gions of the local starburst galaxy Arp220
(shown by blue triangles in Figure 5), al-though they are scaled up by a factor 10 in both size and luminosity (Sakamoto et al. 2008). Thus, the energetics of the star-forming regions within SMMJ2135-0102 are unlike anything found in the present-day Universe, yet the relations between size and luminosity are similar to local, dense GMC cores, suggesting that the underlying physics of the star-forming processes is similar. Overall, these results suggest that the recipes developed to understand star-forming processes in the Milky Way and local galaxies can be used to model the star formation processes in these high-redshift galaxies.
Outlook
Overall, these results provide unique insight into the physics of star formation within a galaxy at z ~ 2 on scales that would otherwise only be achieved with the increased light grasp and resolution of the next generation of facilities, such as ALMA. Indeed, due to the lensing effect we are able to provide an effective glimpse of two of the three science driv-ers for ALMA: to provide images at an angular resolution of 0.1 arcsecond of the dust continuum from distant galaxies; detect spectral CO emission in star-form-ing galaxies out z = 3 and measure their redshifts; albeit only for a single object. Thus, due to the fortuitous discovery with LABOCA and subsequent APEX/SHFI, SMA and GBT/Zpectrometer follow-up of this galaxy, we are able to provide unique insights into the key science that will be routine once ALMA reaches full science operations.
References
Baugh, C. M. et al. 2004, MNRAS, 69, 3101 Caldwell, D. A. et al. 1996, ApJ, 472, 611 Chapman, S. C. et al. 2005, ApJ, 622, 772 Coppin, K. E. K. et al. 2008, MNRAS, 384, 1597 Hill, T. et al. 2005, MNRAS, 363, 405
Lilly, S. et al. 1999, ApJ, 518, 641 Livanou, E. et al. 2006, A&A, 451, 431 Perez-Torres, M. A. et al. 2009, A&A, 507, L17 Sakamoto, N. et al. 2008, ApJ, 684, 957 Scoville, N. et al. 1989, ApJ, 339, 149 Smail, I. et al. 1997, ApJL, 490, 5 Snell, R. L. et al. 2002, ApJ, 578, 229 Swinbank, A. M. et al. 2010, Nature (in press) Tacconi, L. et al. 2008, ApJ, 680, 246 Weiss, A. et al. 2005, A&A, 440, 45 Further images in Press Release eso1012.
